360 J. D. Dana—Glacial Phenomena. 
may have here plunged down—like some rivers in glacial 
Greenland—causing a local deposition of the stones and earth 
that were in the ice, and thus have located and formed the till- 
ade hill. , 
wrenching supposed. : 
The plunging waters would account also for the size and 
position of the half-encircling valley, cut in the schists about 
Rou ill; for this valley’s being mostly free from drift; and 
for the trench which is its continuation eastward to the New 
Haven plain; for this was the way of discharge of the descend- 
ing waters 
But the stream of water descending a crevasse could deposit 
only the drift encountered in its course on and through the ice; 
and wherefore then so high a pile in Round Hill? And why 
were 19-20ths of the drift in the hill the contributions of the 
valley or lower glacier current. 
The height is evidence of long-continued deposition. 
The valley source of the material proves that the drift of the 
glacier was mostly confined to the lower ice. 
he mixture along with the trap and sandstone of some 
porphyritic gneiss, quartzyte, etc., from the northwestward, and 
the increased amount of the same in the western part of the 
hill, shows that the upper ice-stream was supplying material at 
the same time with the lower, though in much smaller amount. 
The height of the hill, in connection with the nature of the 
material, indicates, further, that the lower, or valley, ice-stream 
had a depth in that region of at least 300 feet. 
Round Gill is an example of a “kame.” ‘It is probable 
1 
Conclusions deduced in the preceding pages. 
1. Two movements existed in the glacier-ice—a lower along 
the valley, an upper crossing it obliquely. 
2. The two movements were simultaneous. 
3. The upper ice kept its motion over the southward-flowing 
valley-ice quite across the Connecticut valley in Massachusetts, 
and across ten miles at least of its breadth in the southern half 
of Connecticut. 
